GANAIr: fit a Bayesian model for the integration of nodes,...

In mengyunwu2020/JNN: Joint gene-gene interaction and network selection via a pathway-based Bayesian approach

Description

One of the main functions in the JNNI package.

Usage

GANAIr(X, pathway, L = NULL, t, s2, prob)

Arguments

X	a matrix of predictor variable
pathway	a vector of pathway information, for instance, value 1 represent the corresponding node location belongs to pathway 1
L	a list of Laplace information, each of which is a matrix capturing the relationship between node in the corresponding pathway, default is NULL
t	a vector of the response variable
s2	a numeric
prob	a numeric

Value

A list containing values as follows:

mean_w_red	The coefficients vector
beta_interaction	The coefficients of interaction terms
beta_main	The coefficients of main effects
BIC	Bayesian Information Criterion
df	The number of nonzero coefficients
index_path	Selected networks' number

Examples

library(JNNI)
set.seed(1)
X=matrix(rnorm(50),10,5)
pathway=c(1,1,2,2,3)
b=c(3,3,0,0,0)
t=as.numeric(X%*%b+rnorm(10))
s2=1e-3
prob=0.96
temp=GANAIr(X,pathway,L=NULL,t,s2,prob)

mengyunwu2020/JNN documentation built on June 29, 2020, 12:54 a.m.